C-OBGET (OBGET for Children)

• Psychological dataset for children can be used to train machine learning models which can classify and detect mental disorders in childhood, and prevents the progression of the disorders. • We collected Original Bender Gestalt Drawing Test dataset for children for automating Original Bender Gestalt Drawing Test for children. • There is no psychological drawing test dataset for children with related metadata. To the best of our knowledge most researchers focus on computerized psychological tests for adults, while many mental disorders in adults, initiate childhood problems. Therefore, by detecting mental disorders in childhood, it is possible to prevent many irreparable mental disorders in adulthood. • Collecting children hand sketch dataset is very challenging, children spatial visualization is incomplete and they cannot draw shapes correctly. • Collecting, analyzing, and labeling children's hand sketches are harder than adult hand sketches because it takes more work to explain test rules to children. Children do not sit for a certain period and draw specific shapes; spatial visualization is incomplete. • Children’s OBGET data collection steps and pre-processing are presented. • Children’s OBGET samples are labeled by the proposed semi-automatic labeling.

• 本儿童心理学数据集可用于训练机器学习模型，以分类、检测儿童期精神障碍，并阻断该类障碍的病情进展。
• 我们采集了儿童版本德尔格式塔绘画测验（Original Bender Gestalt Drawing Test）数据集，以实现该测验的自动化实施。
• 目前尚无带有相关元数据的儿童心理绘画测验数据集。据我们所知，多数研究者聚焦于成人计算机化心理测验，而成人的诸多精神障碍往往起源于儿童期的问题。因此，早期检测儿童期精神障碍，可有效预防成年后出现诸多无法逆转的精神疾患。
• 儿童手绘草图数据集的采集极具挑战性，儿童的空间可视化能力尚未发育完全，无法准确绘制各类图形。
• 相较于成人手绘草图，儿童手绘草图的采集、分析与标注难度更高：需耗费更多精力向儿童讲解测验规则，且儿童无法在固定时段内绘制指定图形，空间可视化能力亦存在不足。
• 本文详述了儿童版本德尔格式塔绘画测验（OBGET）的数据采集流程与预处理方案。
• 本研究采用所提出的半自动标注方法，对儿童版本德尔格式塔绘画测验样本进行标注。

搜集汇总

背景与挑战

背景概述

C-OBGET是一个专为儿童心理障碍检测设计的心理学数据集，包含386名4-11岁儿童的手绘图案样本及相关元数据，通过半自动标注支持机器学习模型训练，用于早期发现和预防儿童心理问题。